Model comparison
Magistral Small vs Mistral Large
Mistral Large is the stronger model overall, scoring 31.9 to 30.2 on the Noometry Index. Magistral Small costs 4.0× less per token, which makes it the better buy when Mistral Large's lead doesn't matter for your workload.
Last verified . 6 shared benchmarks.
Mistral AISummary
- They share 6 benchmarks with published results for both. Magistral Small scores higher in 3 categories and Mistral Large in 1 category; 3 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Mistral Large leads 15.8 to 6.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 30% for Magistral Small and 8.5% for Mistral Large.
- Magistral Small is cheaper at $0.50 / $1.50 per million input/output tokens, against $2 / $6 for Mistral Large.
- Mistral Large accepts more context: 131K tokens versus 128K.
Side by side
| Magistral Small | Mistral Large | |
|---|---|---|
| Provider | Mistral AI | Mistral AI |
| Noometry Index | 30.2 | 31.9 |
| Released | 2025-06-10 | 2024-02-26 |
| Weights | Open | Open |
| Context window | 128K | 131K |
| Max output | 40K | 16K |
| Input $ / M tokens | $0.50 | $2 |
| Output $ / M tokens | $1.50 | $6 |
| Results tracked | 10 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Magistral Small leads
Magistral Small: 38.4 (#176), Mistral Large: 34.3 (#240)
| Benchmark | Magistral Small | Mistral Large |
|---|---|---|
| SciCode | 35.2% | 36.2% |
| BigCodeBench Instruct | — | 30% |
| LiveBench Coding | — | 47.1% |
| LMArena Coding | — | 1277 |
| BigCodeBench Complete | — | 38.3% |
| ALE-Bench | — | 264.7 |
| HumanEval+ | — | 62.2% |
| MBPP+ | — | 59.5% |
Agentic & Tool Use Not comparable
Magistral Small: —, Mistral Large: 28.6 (#89)
| Benchmark | Magistral Small | Mistral Large |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 38.4% |
Reasoning Mistral Large leads
Magistral Small: 6.8 (#350), Mistral Large: 15.8 (#310)
| Benchmark | Magistral Small | Mistral Large |
|---|---|---|
| CritPt | 0.3% | 0% |
| DTBench | 61.3% | 65.1% |
| Epoch Capabilities Index | 133.19 | 128.52 |
| ARC-AGI-2 | 0% | — |
| SimpleBench | — | 22.5% |
| Kagi LLM Benchmark | 6.3% | — |
| ARC-AGI-1 | 5% | — |
| Chess Puzzles | 3% | — |
| LiveBench Reasoning | — | 43.5% |
| LMArena Hard Prompts | — | 1257 |
| LiveBench Data Analysis | — | 50.1% |
| LMCA | — | 16.7% |
| ForecastBench | — | 57.1 |
| LiveBench | — | 48.4% |
Math Magistral Small leads
Magistral Small: 26.2 (#261), Mistral Large: 18.2 (#291)
| Benchmark | Magistral Small | Mistral Large |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 30% | 8.5% |
| Omni-MATH | — | 28.1% |
| LiveBench Math | — | 42.5% |
| LMArena Math | — | 1262 |
| MATH Level 5 | — | 50.3% |
| FrontierMath (Feb 2025 set) | — | 0.3% |
Knowledge Too close to call
Magistral Small: 30.9 (#223), Mistral Large: 30.1 (#230)
| Benchmark | Magistral Small | Mistral Large |
|---|---|---|
| GPQA Diamond | 56.1% | 51.3% |
| MMLU-Pro | — | 59.9% |
| Confabulations | — | 21.4% |
| Vectara Hallucination Rate | — | 4.5% |
| GPQA (HELM) | — | 43.5% |
| LMArena Expert | — | 1232 |
| MMLU | — | 80% |
Multilingual Not comparable
Magistral Small: —, Mistral Large: 40.0 (#219)
| Benchmark | Magistral Small | Mistral Large |
|---|---|---|
| LMArena Non-English | — | 1237 |
| LMArena Chinese | — | 1240 |
| LMArena French | — | 1325 |
| LMArena German | — | 1254 |
| LMArena Japanese | — | 1188 |
| LMArena Korean | — | 1202 |
| LMArena Russian | — | 1257 |
| LMArena Spanish | — | 1268 |
Instruction Following Not comparable
Magistral Small: —, Mistral Large: 67.9 (#191)
| Benchmark | Magistral Small | Mistral Large |
|---|---|---|
| LiveBench Instruction Following | — | 67.9% |
| IFEval | — | 87.7% |
| LMArena Instruction Following | — | 1249 |
Long Context Not comparable
Magistral Small: —, Mistral Large: 38.3 (#199)
| Benchmark | Magistral Small | Mistral Large |
|---|---|---|
| LMArena Longer Query | — | 1261 |
Writing & Preference Not comparable
Magistral Small: —, Mistral Large: 40.7 (#242)
| Benchmark | Magistral Small | Mistral Large |
|---|---|---|
| LMArena Text | — | 1266 |
| LMArena Creative Writing | — | 1243 |
| Short-Story Creative Writing | — | 69% |
| EQ-Bench Creative Writing | — | 985 |
| WildBench | — | 80.1% |
| LMArena Multi-Turn | — | 1260 |
| LiveBench Language | — | 39.4% |
Frequently asked questions
Is Magistral Small better than Mistral Large?
Mistral Large is the stronger model overall, scoring 31.9 to 30.2 on the Noometry Index. Magistral Small costs 4.0× less per token, which makes it the better buy when Mistral Large's lead doesn't matter for your workload.
Which is cheaper, Magistral Small or Mistral Large?
Magistral Small is cheaper. It lists at $0.50 per million input tokens and $1.50 per million output tokens; Mistral Large lists at $2 and $6.
Is Magistral Small or Mistral Large better for coding?
Magistral Small scores higher on coding benchmarks: 38.4 versus 34.3 in the Noometry coding category.
Which has the bigger context window?
Mistral Large does, with 131K tokens against 128K.
How many benchmarks do Magistral Small and Mistral Large share?
6 benchmarks have published results for both models. Magistral Small has 10 scored results on Noometry and Mistral Large has 51.